Multi-user mixed multi-hop relay network

Abstract

The present disclosure relates to a triple-hop multiuser relay network. Further, the relay network is comprised of mixed communication mediums (radiofrequency/free-space optical/radiofrequency), and utilizes a generalized order user scheduling scheme for determining the next source or destination to be selected for transmission. Closed-form expressions were achieved to describe outage probability, average symbol error probability, and channel capacity assuming Rayleigh and Gamma-Gamma fading models for the radiofrequency and free-space optical links, respectively. The effects of pointing errors on the free-space optical link were also considered. Additionally, a power allocation algorithm was proposed to optimize power allocation at each hop.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

The present application claims priority to U.S. Provisional Application No. 62/578,090, filed Oct. 27, 2017, the teaching of which is hereby incorporated by reference in its entirety for all purposes.

STATEMENT REGARDING PRIOR DISCLOSURE BY THE INVENTORS

Aspects of this technology are described in an article “A new scenario of triple-hop mixed RF/FSO/RF relay network with generalized order user scheduling and power allocation” published in EURASIP Journal on Wireless Communications and Networking, on Oct. 28, 2016, which is incorporated herein by reference in its entirety.

BACKGROUND

Field of the Disclosure

The present disclosure relates to a multi-hop relay network with generalized order user scheduling and transmission power allocation.

Description of the Related Art

Cooperative relay networks present an efficient solution for multipath fading issues in wireless communications. In these relay networks, a relay node or a set of relay nodes facilitate propagation of a message from one or more source nodes to one or more destination nodes, thereby providing diversity, widening coverage area, and reducing the need for high power transmitters. Pursuant to network architecture and desired functionality, each node may employ either an amplify-and-forward (AF) or decode-and-forward (DF) scheme. The simplicity of AF schemes is often weighed against the computationally demanding, but improved output, of DF schemes.

Recent efforts to reduce power consumption, expand coverage and improve reliability of wireless communications have employed a mixture of data transmission modalities. By employing relays and varied transmission modalities, networks are able to increase communication distance and improve network diversity. To this end, several approaches have been explored, including single-relay free space optical (FSO) communications and dual-hop mixed radiofrequency (RF) and FSO relay networks. Further, by considering multiple users a network aims to achieve multiuser diversity.

While recent work has considered triple-hop relaying for only one type of transmission modality, broad advances have focused on dual-hop mixed RF/FSO relay networks, representative of applications where multiple users communicate with a relay node via RF links and the relay forwards their messages to a base station over an FSO link.

Therefore, one objective of the present disclosure is to provide and evaluate a model of triple-hop mixed-mode relaying in wireless networks.

The foregoing “Background” description is for the purpose of generally presenting the context of the disclosure. Work of the inventors, to the extent it is described in this background section, as well as aspects of the description which may not otherwise qualify as prior art at the time of tiling, are neither expressly or impliedly admitted as prior art against the present invention.

SUMMARY

The present disclosure relates to a triple-hop mixed RF/FSO/RF relay network with generalized order user scheduling and transmission power allocation.

The triple-hop mixed RF/FSO/RF relay network includes K₁ sources, two DF relays, and K₂ destinations. The sources and destinations are connected to respective relay nodes via RF links, while the relay nodes are connected via FSO link.

The generalized order user scheduling scheme selects the source with the N₁^th best signal-to-noise ratio (SNR) among the available sources to communicate with the first relay node. Similarly, following transmission via FSO from the first relay node to the second relay node, the destination with the N₂^th best SNR is selected to receive its message from the second relay.

Further, optimum transmission powers of the selected user are obtained on the first hop, first relay, and second relay.

The foregoing paragraphs have been provided by way of general introduction, and are not intended to limit the scope of the following claims. The described embodiments, together with further advantages, will be best understood by reference to the following detailed description taken in conjunction with the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

A more complete appreciation of the disclosure and many of the attendant advantages thereof will be readily obtained as the same becomes better understood by reference to the following detailed description when considered in connection with the accompanying drawings, wherein:

FIG. 1 is a schematic of a multiuser multi-hop mixed transmission modality network with generalized order user scheduling according to one or more aspects of the disclosed subject matter;

FIG. 2 is a schematic of a multiuser triple-hop mixed RF/FSO/RF relay network with generalized order user scheduling according to one or more aspects of the disclosed subject matter;

FIG. 3 is a graphical representation of system outage probability versus SNR of a multiuser mixed RF/FSO/RF relay network with generalized order user scheduling for different values of N₁=N₂ according to one or more aspects of the disclosed subject matter;

FIG. 4 is a graphical representation of system outage probability versus SNR of a multiuser mixed RF/FSO/RF relay network with generalized order user scheduling for different values of K₁=K₂ according to one or more aspects of the disclosed subject matter;

FIG. 5 is a graphical representation of system outage probability versus SNR of a multiuser mixed RF/FSO/RF relay network with generalized order user scheduling for different values of γ_out with fixed average SNRs and varying average SNRs according to one or more aspects of the disclosed subject matter;

FIG. 6 is a graphical representation of system outage probability versus order of selected user of multiuser mixed RF/FSO/RF relay network with generalized order user scheduling for different values of average SNR according to one or more aspects of the disclosed subject matter;

FIG. 7 is a graphical representation of system outage probability versus SNR of multiuser mixed RF/FSO/RF relay network with generalized order user scheduling for different values of γ_out with and without power optimization according to one or more aspects of the disclosed subject matter;

FIG. 8 is a graphical representation of system average symbol error probability versus SNR of a multiuser mixed RF/FSO/RF relay network with generalized order user scheduling for different values of ζ according to one or more aspects of the disclosed subject matter;

FIG. 9 is a graphical representation of system average symbol error probability versus SNR of a multiuser mixed RF/FSO/RF relay network with generalized order user scheduling for different values of α, β, and r according to one or more aspects of the disclosed subject matter;

FIG. 10 is a graphical representation of system average symbol error probability versus SNR of a multiuser mixed RF/FSO/RF relay network with generalized order user scheduling for different values of (K₁, N₁) and (K₂, N₂) according to one or more aspects of the disclosed subject matter;

FIG. 11 is a graphical representation of ergodic capacity versus SNR of a multiuser mixed RF/FSO/RF relay network with generalized order user scheduling for different values of N₁=N₂ according to one or more aspects of the disclosed subject matter; and

FIG. 12 is a hardware block diagram of a server according to one or more exemplary aspects of the disclosed subject matter.

DETAILED DESCRIPTION

The terms “a” or “an”, as used herein, are defined as one or more than one. The term “plurality”, as used herein, is defined as two or more than two. The term “another”, as used herein, is defined as at least a second or more. The terms “including” and/or “having”, as used herein, are defined as comprising (i.e., open language). Reference throughout this document to “one embodiment”, “certain embodiments”, “an embodiment”, “an implementation”, “an example” or similar terms means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present disclosure. Thus, the appearances of such phrases or in various places throughout this specification are not necessarily all referring to the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments without limitation.

Recently, incorporation of free-space optical communication in relay networks has been proposed to provide improved reliability in the ‘last-mile’ of wireless communications. These networks, often known as dual-hop relay networks, transmit a source message from a source to a relay node over a RF link (licensed frequencies) and then forward the message to the destination over an FSO link (license-free). In such networks, relays provide greater diversity among nodes, expand the coverage area, and reduce the need for high-power transmitters. Further, this approach can be complemented via multiuser cooperation and opportunistic scheduling.

Extensive work has been dedicated to the above, filling the connectivity gap in ‘last-mile’ connectivity while conserving economic resources and saving bandwidth by exploiting optical communications. To this point, however, FSO communications have been incorporated into dual-hop networks with the FSO link connecting directly to a base station at a terminus. Therefore, a triple-hop mixed relay network with generalized user scheduling and power allocation algorithm, yet to be developed, is described in the present disclosure.

FIG. 1 shows a triple-hop mixed RF/FSO/RF relay network consisting of K₁ sources on a first hop 117 U_k (k=1, 2, . . . , K₁) (102, 103, 104), two un-coded type DF relays R_i (i=1, 2) (105, 115) and K₂ destinations on a third hop 119 D_j (j=1, 2, . . . , K₂) (112, 113, 114). The sources 110 are connected with a first relay 105 through RF links 117. The first relay 105 is connected with a second relay 115 via FSO link. The second relay 115 is connected with destinations 120 through RF links 119. The abovementioned communication modalities are non-limiting and are merely representative of a variety of communication modalities. The direct links between the sources and destinations are assumed to be in deep fade. Further, channel coefficients are held constant over an entire block of communication in accordance with a block fading model.

Communication is operated in a half-duplex mode and to be conducted over three phases outlined above and repeated here: selected user U_set→R₁, R₁→R₂, and R₂→D_set. The received signal at R₁ 105 from the kth user can be expressed as
y_k,r1=/√{square root over (P_k)}h_k,r1x_k,r1+n_r1,  (1)
where P_k is the transmit power of the kth user, h_k,r1 is the channel coefficient of the U_k→R₁ link, x_k,r1 is the transmitted symbol of U_k with {|x_k,r1|²}=1, and n_r1˜N(0, N₀₁) is an additive white Gaussian noise (AWGN) term, where {⋅} is the mathematical expectation. Using (1), the SNR at R₁ 105 due to U_k 104 can be written as

$\begin{array}{cc}\gamma U_k,R_1=\frac{P_k}{N_{01}}{h_{k,r_1}}^2.& \left(2\right)\end{array}$

According to the generalized order user scheduling, the source with the N₁^th best γ_Uk,R1 or equivalently, the N₁^th largest |h_k,r1|² among the other sources 110 is selected to transmit its message to R₁ 105 in the first communication phase 117. In other words, the source 110 is selected such that γ_USelR1=N₁^th max{γ_Uk,R1}. Processing circuitry is configured to perform the selection process at R₁ according the selection scheme described herein. The subcarrier intensity modulation (SIM) scheme is employed at the relay R₁ 105, where a standard RF coherent/noncoherent modulator and demodulator can be used for transmitting and recovering the source data. At R₁ 105, after filtering by a bandpass filter (BPF), a direct current (DC) bias is added to the filtered RF signal to ensure that the optical signal is non-negative. Then, the biased signal is sent to a continuous wave laser driver. The retransmitted optical signal at R₁ 105 is written as
y_r1^Opt=√{square root over (P_Opt)}(1+y_Sel,r1),  (3)
where P_Opt denotes the average transmitted optical power and it is related to the relay electrical power P_r by the electrical-to-optical conversion efficiency η₁ as P_Opt=η₁ P_r1, where M denotes the modulation index and γ_Sel,r1 is the RF received signal at R₁ 105 from the selected source (see article by Lee, E et al, “Performance analysis of the asymmetric dual-hop relay transmission with mixed RF/FSO links” published in IEEE Transactions on Information Theory, in 2004, and incorporated herein by reference). The optical signal at R₂ 115 received from R₁ 105 at the second phase of communication 118 can be expressed as
γ_r1,r2=g_r1,r2{√{square root over (P_Opt)}[1+(√{square root over (P_Sel)}h_Sel,r1x_Sel,r1+n_r1)]}+n_r2,  (4)
where n_r2˜N(0, N₀₂) is an AWGN term at R₂ 115. Moreover, the channel coefficients of the R₁→R₂ link which is given by g_r1,r2 is modeled as g_r1,r2=g_ag_f, where g_a and g_f are the average gain and the fading gain of the FSO link, respectively, and are given by

$\begin{array}{cc}\{\begin{array}{c}{g}_a={\left[\mathrm{erf}\left(\frac{\sqrt{\pi }q}{\sqrt{2}\varphi d^{\mathrm{FSO}}}\right)\right]}^2\times {10}^{-\kappa d^{\mathrm{FSO}}/10},\\ g_f\sim \mathrm{GGamma}\left(\alpha ,\beta \right),\end{array}& \left(5\right)\end{array}$
where q is the aperture radius, ϕ is the divergence angle of the beam, d^FSO is the distance between the FSO transmitter and receiver, κ is the weather-dependent attenuation coefficient, and GGamma(α, β) represents a Gamma-Gamma random variable with parameters α and β (see article by Zhang, W et al, “Soft-switching hybrid FSO/RF links using short-length raptor codes: design and implementation” published in IEEE Journal on Selected Areas in Communications, in 2009, and incorporated herein by reference). Assuming spherical wave propagation, the parameters α and β in the Gamma-Gamma distribution, which represent the fading turbulence conditions, are related to the physical parameters as follows:

$\begin{array}{cc}\alpha ={\left[\exp \left\{\frac{0.49{\vartheta }^2}{{\left[1+0.18{\xi }^2+0.56{\vartheta }^{12/5}\right]}^{7/6}}\right\}-1\right]}^{-1},& \left(6\right)\\ \beta ={\left[\exp \left\{\frac{0.51{{\vartheta }^2\left[1+0.69{\vartheta }^{12/5}\right]}^{-5/6}}{{\left[1+0.9{\xi }^2+0.62{\xi }^2{\vartheta }^{12/5}\right]}^{5/6}}\right\}-1\right]}^{-1},& \left(7\right)\end{array}$
where ϑ²=0.5 C_n²ζ^7/6(d^FSO)^11/6, ξ²=ζq²/d^FSO, ζ=2π/λ^FSO is the wavelength, and C_n² is the weather-dependent index of refraction structure parameter (see article by He, B et al, “Bit-interleaved coded modulation for hybrid RF/FSO systems” published in IEEE Transactions on Communications, in 2009, and incorporated herein by reference).

When the DC component is filtered out at R₂ 115 and an optical-to-electrical conversion is performed, assuming M=1, the received signal can be expressed as follows:
γ_r1,r2=g_r1,r2√{square root over (P_Ele)}(√{square root over (P_Sel)}h_Sel,r1x_Sel,r2+n_r1)+n_r2,  (8)
where P_Ele=η₂P_Opt=η₁η₂P_r1 is the electrical power received at R₂ 115 and η₂ is the optical-to-electrical conversion efficiency.

From (8), the SNR at R₂ 115 can be written as

$\begin{array}{cc}\gamma R_2=\frac{\gamma U_{\mathrm{Sel}},R_1\gamma R_1,R_2}{\gamma \mathrm{Sel},R_1+\gamma R_1,R_2+1},\mathrm{where}{\gamma }_{U_{\mathrm{Sel}},R_1}=\frac{P_{\mathrm{Sel}}}{N_{01}}{h_{\mathrm{Sel},r_1}}^2,{\gamma }_{R_1,R_2}=\frac{{\eta }_1{\eta }_2{P}_{r_1}}{N_{02}}{g_{r_1,r_2}}^2,& \left(9\right)\end{array}$
and P_r1 is the transmit power at R₁ 105. The SNR in (9) can be rewritten using the standard approximation γ_R2 min(γ_USel,R1,γ_R1,R2) (see article by Ansari, I S et al, “Impact of point errors on the performance of mixed RF/FSO dual-hop transmission systems” published in IEEE Wireless Communications Letters, in 2013 and an article by Ansari, I S et al, “On the performance of mixed RF/FSO variable gain dual-hop transmission systems with pointing errors”, published at IEEE Vehicular Technology Conference, in 2013, and incorporated herein by reference) as

$\begin{array}{cc}\gamma R_2=\frac{\gamma U_{\mathrm{Sel}},R_1\gamma R_1,R_2}{\gamma U\mathrm{Sel},R_1+\gamma R_1,R_2+1}.& \left(10\right)\end{array}$

The signal received at D_j 114 from R₂ 115 in the third phase of communication 119 can be written as
γ_r2,dj=/√{square root over (P_r2)}h_r2,jx_dj+n_dj,  (11)
where P_r2 is the transmit power at R₂ 115, h_r2,j is the channel coefficient of the R₂→D_j link, x_dj is the transmitted symbol of d_j with {|x_dj|²}=1, and n_dj˜N(0, N₀₃) is an AWGN term. Using (11), the SNR at D_j 114 can be written as

$\begin{array}{cc}{\gamma }_{R_2,D_j}=\frac{P_{r_2}}{N_{03}}{h_{r_2,j}}^2.& \left(12\right)\end{array}$

According to generalized order user scheduling, the destination 120 with the N₂^th best γ_R2,Dj or equivalently, the N₂^th largest |h_r2,j|² among the other destinations 120 is selected to receive its message from R₂ 115 in the third communication phase 119. In other words, the destination is selected such that γ_R2,DSel=N₂^th max{γ_R2,Dj}. Processing circuitry is configured to perform the selection process at R₂ according the selection scheme described herein.

The channel coefficients of the RF links 117, 119 h_k,r1 (k=1= . . . =K₁) and h_r2,j (j=1= . . . =K₂) follow the Rayleigh fading model and, therefore, channel gains |h_k,r1|² and |h_r2,j|² are exponentially distributed random variables with mean powers Ω_k,r1 and Ω_r2,j, respectively. Therefore, the probability density functions (PDFs) of γ_Uk,R1, and γ_R2,Dj are given by

${\gamma U_k,R_1}_{}\left(\gamma \right)={\lambda }_{k,r_1}\exp \left(-{\lambda }_{k,r_1}\gamma \right),\mathrm{where}{\lambda }_{k,r_1}=1/{\stackrel{\_}{\gamma }}_{k,r_1}\mathrm{and}$ ${\stackrel{\_}{\gamma }}_{k,r_1}=\frac{P_k}{N_0}𝔼\left\{{h_{k,r_1}}^2\right\}=\frac{P_k}{N_{01}}{\Omega }_{k,r_1},\mathrm{and}\mathrm{by}$ ${\gamma R_2,D_j}_{}\left(\gamma \right)={\lambda }_{r_2,j}\exp \left(-{\lambda }_{r_2,j}\gamma \right),\mathrm{where}$ ${\lambda }_{r_2,j}=1/{\stackrel{\_}{\gamma }}_{r_2,j}\mathrm{and}{\stackrel{\_}{\gamma }}_{r_2,j}=\frac{P_k}{N_{03}}𝔼\left\{{h_{r_2,j}}^2\right\}=\frac{P_{r_2}}{N_{03}}{\Omega }_{r_2,j},$
respectively. Regarding the second hop, it is assumed that the FSO link 118 experiences a unified Gamma-Gamma fading model including the pointing errors effect whose SNR PDF (see article by Ansari, I S et al, “Impact of point errors on the performance of mixed RF/FSO dual-hop transmission systems” published in IEEE Wireless Communications Letters, in 2013, and incorporated herein by reference), is given by

$\begin{array}{cc}{f}_{{\gamma }_{R_1,R_2}}\left(\gamma \right)=\frac{{\zeta }^2}{r\mathrm{\gamma \Gamma }\left(\alpha \right)\Gamma \left(\beta \right)}{G}_{1,3}^{3,0}\left[\alpha {\beta \left({\lambda }_{r_1,r_2}\gamma \right)}^{\frac{1}{r}}❘\begin{array}{c}{\zeta }^2+1\\ {\zeta }^2,\alpha ,\beta \end{array}\right],& \left(13\right)\end{array}$
ζ is the ratio between the equivalent beam radius at the receiver and the pointing error displacement standard deviation (jitter) at the receiver (i.e. when ζ→∞, non-pointing error). r is the parameter defining the type of detection technique (i.e. r=1 represents heterodyne detection and r=2 represents intensity modulation (IM)/direct detection (DD)). α and β are the fading parameters related to the atmospheric turbulence conditions with lower values indicating severe atmospheric turbulence conditions. Γ(.) is the Gamma function,

${\lambda }_{r_1,r_2}=1/{\stackrel{\_}{\gamma }}_{r_1,r_2}\mathrm{where}$ ${\stackrel{\_}{\gamma }}_{r_1,r_2}=\frac{{\eta }_1{\eta }_2{P}_{r_1}}{N_{02}}𝔼\left\{{g_{r_1,r_2}}^2\right\}=\frac{{\eta }_1{\eta }_2{P}_{r_1}}{N_{02}}{\mu }_{r_1,r_2}$
and G(.) is the Meijer G-function (see textbook by Gradshteyn, I S and Ryzhik, I M, “Tables of Integrals, Series and Products”, published by Academic Press, in 2000, and incorporated herein by reference).

The end-to-end (e2e) SNR at the selected destination can be written using the standard approximation γ_D min(γ_R2, γ_R2,DSel) as

$\begin{array}{cc}{\gamma }_D=\frac{{\gamma }_{R_2}{\gamma }_{R_2,D_{\mathrm{Sel}}}}{{\gamma }_{R_2}+{\gamma }_{R_2,D_{\mathrm{Sel}}}}.& \left(14\right)\end{array}$

In an exemplary embodiment, as seen in FIG. 2, K₁ sources 210 are users of wireless devices. A user U_Sel 204 is selected from a group of source users 210, as the user with the N₁^th best SNR value, to transmit its signal to a first relay 201. The signal is transmitted via RF link on a first hop 217. Each source user (U₁ 202, U₂ 203, . . . , U_K1 204) is equipped with a single antenna. The first relay 201 is equipped with a single antenna 205 and a single photo-aperture transmitter 206. A second relay 211 is equipped with a single antenna 215 and a photo-aperture transmitter 216. Deploying a DF scheme, the first relay 201, with support from processing circuitry configured to decode a transmitted signal, receives the signal from the user U_Sel 204 and transmits the signal to a second relay 211 on a second hop 218 via FSO link. The second relay 211 receives the signal from the first relay 201 and, deploying a DF scheme, with support from processing circuitry configured to decode the transmitted signal, transmits the signal to a destination user D_Sel 214 via RF link on a third hop 219. The destination user D_Sel 214 is selected from a group of destination users 220, as the user with the N₂^th best SNR value, to receive its signal from the second relay 211. Each destination user (D₁ 212, D₂ 213, . . . , D_K2 214) is equipped with a single antenna.

System Performance Metrics (Analytical Solutions)

To evaluate system performance, the statistics of the e2e SNR provided in (14) must be determined.

To this end, the outage probability is defined as the probability that the SNR at a selected destination drops below a predetermined outage threshold γ_out, or P_out=Pr[γ_d≤γ_out], where Pr[.] is the probability operation and γ_out is a predetermined outage threshold. The outage probability can be obtained from the cumulative distribution function (CDF) of the e2e SNR as P_out=F_γD(γ_out). This CDF can be written in terms of CDFs of the three hops' SNRS as

$\begin{array}{cc}{F}_{{\gamma }_D}\left(\gamma \right)=1-\left\{\left(1-F_{{\gamma }_{U_{\mathrm{Sel}},R_1}}\left(\gamma \right)\right)\left(1-F_{{\gamma }_{R_1,R_2}}\left(\gamma \right)\right)\times \left(1-F_{{\gamma }_{R_2,D_{\mathrm{Sel}}}}\left(\gamma \right)\right)\right\},\mathrm{where}{F}_{{\gamma }_{U_{\mathrm{Sel}},R_1}}\left(\gamma \right),F_{{\gamma }_{R_1,R_2}}\left(\gamma \right),F_{{\gamma }_{R_2,D_{\mathrm{Sel}}}}\left(\gamma \right)& \left(15\right)\end{array}$
are the CDFs of the first hop, second hop, and third hop SNRs, respectively.

The CDF of the first hop begins from the PDF according to generalized order user selection, wherein the PDF represents the N₁^th best SNR or, the source of the N₁^th best SNR as selected by the first relay. The CDF is rewritten as

$\begin{array}{cc}{F}_{{\gamma }_{U_{\mathrm{Sel}},R_1}}\left(\gamma \right)=K_1\left(\begin{array}{c}{K}_1-1\\ N_1-1\end{array}\right)\sum _{k=0}^{K_1-N_1}\frac{\left(\begin{array}{c}{K}_1-N_1\\ k\end{array}\right){\left(-1\right)}^k}{\left(k+N_1\right)}\times \left[1-\exp \left(-\left(k+N_1\right){\lambda }_{u,r_1}\gamma \right)\right].& \left(19\right)\end{array}$
where the users on the third hop have been assumed to have independent identical distributed channels.

The CDF of the second hop is determined from the PDF of the FSO link incorporating the Gamma-Gamma fading model and including point errors. The CDF is rewritten as

$\begin{array}{cc}{F}_{{\gamma }_{R_1,R_2}}\left(\gamma \right)={\mathrm{AG}}_{r+1,3r+1}^{3r,1}\left[\frac{B}{{\stackrel{\_}{\gamma }}_{r_1,r_2}}\gamma ❘\begin{array}{c}1,{\chi }_1\\ {\chi }_2,0\end{array}\right],\mathrm{where}A=\frac{r^{\alpha +\beta -2\mathrm{\zeta 2}}}{{\left(2\pi \right)}^{r-1}\Gamma \left(\alpha \right)\Gamma \left(\beta \right)},B=\frac{{\left(\mathrm{\alpha \beta }\right)}^r}{r^{2r}},{\chi }_1=\frac{{\zeta }^2+1}{r},...,\frac{{\zeta }^2+r}{r},& \left(20\right)\end{array}$
comprises of r terms and

${\chi }_2=\frac{{\zeta }^2}{r},...,\frac{{\zeta }^2+r-1}{r},\frac{\alpha }{r},...,\frac{\alpha +r-1}{r},\frac{\beta }{r},...,\frac{\beta +r-1}{r}$
comprises of 3r terms.

Similar to the first hop, the CDF of the third hop begins from the PDF according to generalized order user selection, wherein the PDF represents the N₂^th best SNR or, the destination of the N₂^th best SNR as selected by the second relay. The CDF is rewritten as

$\begin{array}{cc}{F}_{{\gamma }_{R_2,D_{\mathrm{Sel}}}}\left(\gamma \right)=K_2\left(\begin{array}{c}{K}_2-1\\ N_2-1\end{array}\right)\sum _{j=0}^{K_2-N_2}\frac{\left(\begin{array}{c}{K}_2-N_2\\ j\end{array}\right){\left(-1\right)}^j}{\left(j+N_2\right)}\times \left[1-\exp \left(-\left(j+N_2\right){\lambda }_{r2,u}\gamma \right)\right],& \left(22\right)\end{array}$
where the users on the third hop have been assumed to have independent identical distributed channels.

Following the substitution of the CDFs from each hop ((19), (20), (22)) into (15), the full e2e CDF can be written as

$\begin{array}{cc}{F}_{{\gamma }_D}\left(\gamma \right)=\left(\begin{array}{c}{K}_1-1\\ N_1-1\end{array}\right)K_1\sum _{k=0}^{K_1-N_1}\frac{\left(\begin{array}{c}{K}_1-N_1\\ k\end{array}\right){\left(-1\right)}^k}{\left(k+N_1\right)}\left\{1-\exp \left({\tau }_1\gamma \right)-A\times \left(G_{r+1,3r+1}^{3r,1}\left[{\delta }_0\gamma ❘\begin{array}{c}1,{\chi }_1\\ {\chi }_2,0\end{array}\right]\left[1-\exp \left(-{\tau }_1\gamma \right)\right]\right)\right\}+\left(\begin{array}{c}{K}_2-1\\ N_2-1\end{array}\right)K_2\sum _{j=0}^{K_2-N_2}\frac{\left(\begin{array}{c}{K}_2-N_2\\ j\end{array}\right){\left(-1\right)}^j}{\left(j+N_2\right)}\left\{1-\exp \left(-{\tau }_2\gamma \right)-A\left(G_{r+1,3r+1}^{3r,1}\left[{\delta }_0\gamma ❘\begin{array}{c}1,{\chi }_1\\ {\chi }_2,0\end{array}\right]\left[1-\exp \left(-{\tau }_2\gamma \right)\right]\right)\right\}-\left(\begin{array}{c}{K}_1-1\\ N_1-1\end{array}\right)K_1\sum _{k=0}^{K_1-N_1}\frac{\left(\begin{array}{c}{K}_1-N_1\\ k\end{array}\right){\left(-1\right)}^k}{\left(k+N_1\right)}\left(\begin{array}{c}{K}_2-1\\ N_2-1\end{array}\right)K_2\times \hspace{1em}\sum _{j=0}^{K_2-N_2}\frac{\left(\begin{array}{c}{K}_2-N_2\\ j\end{array}\right){\left(-1\right)}^j}{\left(j+N_2\right)}\left\{1-\exp \left(-{\tau }_1\gamma \right)-\exp \left(-{\tau }_2\gamma \right)+\exp \left(-\left[{\tau }_1+{\tau }_2\right]\gamma \right)-A\left(G_{r+1,3r+1}^{3r,1}\left[{\delta }_0\gamma ❘\begin{array}{c}1,{\chi }_1\\ {\chi }_2,0\end{array}\right]\times \left[1-\exp \left(-{\tau }_1\gamma \right)-\exp \left(-{\tau }_2\gamma \right)+\exp \left(-\left[{\tau }_1+{\tau }_2\right]\gamma \right)\right]\right)\right\}+{\mathrm{AG}}_{r+1,3r+1}^{3r,1}\left[{\delta }_0\gamma ❘\begin{array}{c}1,{\chi }_1\\ {\chi }_2,0\end{array}\right],\mathrm{where}{\tau }_1=\left(k+N_1\right){\lambda }_{u,r_1},{\delta }_0=\frac{B}{{\stackrel{\_}{\gamma }}_{r_1,r_2}},\mathrm{and}{\tau }_2=\left(j+N_2\right){\lambda }_{r_2,u}.& \left(23\right)\end{array}$
The CDF in (23) is used to determine several performance measures as closed-form expressions.

To determine the exact average symbol error probability (ASEP), the ASEP is expressed in terms of the CDF of γ_D as

$\begin{array}{cc}\mathrm{ASEP}=\frac{a\sqrt{b}}{2\sqrt{\pi }}{\int }_0^{\infty }\frac{\exp \left(-b\gamma \right)}{\sqrt{\gamma }}{F}_{{\gamma }_D}\left(\gamma \right)d\gamma ,& \left(24\right)\end{array}$
where a and b are modulation-specific parameters (see article by McKay, M R et al, “Performance analysis of MIMO-MRC in double-correlated Rayleigh environments”, published in IEEE Transactions on Communications, in 2007, incorporated herein by reference). A SIM scheme is adopted, allowing known digital modulation techniques such as phase shift keying to be used. Therefore, the error probability computing method, used for RF wireless communication systems, can be used to evaluate the error probability performance in FSO systems. Upon combination of equations, ASEP can be written as

$\begin{array}{cc}\mathrm{ASEP}=\frac{a\sqrt{b}}{2\sqrt{\pi }}\left\{\left(\begin{array}{c}{K}_1-1\\ N_1-1\end{array}\right)K_1\sum _{k=0}^{K_1-N_1}\frac{\left(\begin{array}{c}{K}_1-N_1\\ k\end{array}\right){\left(-1\right)}^k}{\left(k+N_1\right)}\left(\frac{\Gamma \left(1/2\right)}{b^{\frac{1}{2}}}-\frac{\Gamma \left(1/2\right)}{{\left(b+{\tau }_1\right)}^{\frac{1}{2}}}-A\left[b^{-\frac{1}{2}}{G}_{r+2,3r+1}^{3r,2}\left[\frac{{\delta }_0}{b}❘\begin{array}{c}\frac{1}{2},1,{\chi }_1\\ {\chi }_2,0\end{array}\right]-{\left(b+{\tau }_1\right)}^{-\frac{1}{2}}{G}_{r+2,3r+1}^{3r,2}\left[\frac{{\delta }_1}{\left(b+{\tau }_1\right)}❘\begin{array}{c}\frac{1}{2},1,{\chi }_1\\ {\chi }_2,0\end{array}\right]\right]\right)+\left(\begin{array}{c}{K}_2-1\\ N_2-1\end{array}\right)\times K_2\sum _{j=0}^{K_2-N_2}\frac{\left(\begin{array}{c}{K}_2-N_2\\ j\end{array}\right){\left(-1\right)}^j}{\left(j+N_2\right)}\left(\frac{\Gamma \left(1/2\right)}{b^{\frac{1}{2}}}-\frac{\Gamma \left(1/2\right)}{{\left(b+{\tau }_2\right)}^{\frac{1}{2}}}-A\left[b^{-\frac{1}{2}}{G}_{r+2,3r+1}^{3r,2}\left[\frac{{\delta }_0}{b}❘\begin{array}{c}\frac{1}{2},1,{\chi }_1\\ {\chi }_2,0\end{array}\right]-{\left(b+{\tau }_2\right)}^{-\frac{1}{2}}{G}_{r+2,3r+1}^{3r,2}\times \left[\frac{{\delta }_0}{\left(b+{\tau }_2\right)}❘\begin{array}{c}\frac{1}{2},1,{\chi }_1\\ {\chi }_2,0\end{array}\right]\right]\right)-\left(\begin{array}{c}{K}_1-1\\ N_1-1\end{array}\right)K_1\sum _{k=0}^{K_1-N_1}\frac{\left(\begin{array}{c}{K}_1-N_1\\ k\end{array}\right)}{\left(k+N_1\right)}\times {\left(-1\right)}^k\left(\begin{array}{c}{K}_2-1\\ N_2-1\end{array}\right)K_2\sum _{j=0}^{K_2-N_2}\frac{\left(\begin{array}{c}{K}_2-N_2\\ j\end{array}\right){\left(-1\right)}^j}{\left(j+N_2\right)}\left(\frac{\Gamma \left(1/2\right)}{b^{\frac{1}{2}}}-\frac{\Gamma \left(1/2\right)}{{\left(b+{\tau }_1\right)}^{\frac{1}{2}}}-\frac{\Gamma \left(1/2\right)}{{\left(b+{\tau }_2\right)}^{\frac{1}{2}}}+\frac{\Gamma \left(1/2\right)}{{\left(b+{\tau }_1+{\tau }_2\right)}^{\frac{1}{2}}}-A\left[b^{-\frac{1}{2}}\times G_{r+2,3r+1}^{3r,2}\left[\frac{{\delta }_0}{b}❘\begin{array}{c}\frac{1}{2},1,{\chi }_1\\ {\chi }_2,0\end{array}\right]-{\left(b+{\tau }_1\right)}^{-\frac{1}{2}}{G}_{r+2,3r+1}^{3r,2}\times \left[\frac{{\delta }_0}{\left(b+{\tau }_1\right)}❘\begin{array}{c}\frac{1}{2},1,{\chi }_1\\ {\chi }_2,0\end{array}\right]-G_{r+2,3r+1}^{3r,2}\left[\frac{{\delta }_0}{\left(b+{\tau }_2\right)}❘\begin{array}{c}\frac{1}{2},1,{\chi }_1\\ {\chi }_2,0\end{array}\right]\times {\left(b\times {\tau }_2\right)}^{-\frac{1}{2}}+G_{r+2,3r+1}^{3r,2}\left[\frac{{\delta }_0}{\left(b+{\tau }_1+{\tau }_2\right)}❘\begin{array}{c}\frac{1}{2},1,{\chi }_1\\ {\chi }_2,0\end{array}\right]\times {\left(b+{\tau }_1+{\tau }_2\right)}^{-\frac{1}{2}}\right]\right)+{\mathrm{Ab}}^{-\frac{1}{2}}{G}_{r+2,3r+1}^{3r,2}\left[\frac{{\delta }_0}{b}❘\begin{array}{c}\frac{1}{2},1,{\chi }_1\\ {\chi }_2,0\end{array}\right]\right\}.& \left(25\right)\end{array}$

Because the coherence time of the FSO fading channel is in the order of milliseconds, a single fade can obliterate millions of bits at gigabits/second data rates. Therefore, the exact average (i.e., ergodic) channel capacity represents the best achievable capacity of an optical wireless link. Using a PDF-based method, the ergodic capacity can be expressed in terms of the PDF of γ_D as

$\begin{array}{cc}C=\frac{1}{\ln \left(2\right)}{\int }_0^{\infty }\ln \left(1+\gamma \right)f_{{\gamma }_D}\left(\gamma \right)d\gamma .& \left(26\right)\end{array}$

Following multiple derivations and integrations described in detail in the cited references, the exact ergodic capacity can be written as

$\begin{array}{cc}C=\frac{1}{\ln \left(2\right)}\{\left(\begin{array}{c}{K}_1-1\\ N_1-1\end{array}\right)K_1\sum _{k=0}^{K_1-N_1}\frac{\left(\begin{array}{c}{K}_1-N_1\\ k\end{array}\right){\left(-1\right)}^k}{{\tau }_1}(-\exp \left({\tau }_1\right)E_i\left(-{\tau }_1\right)-\hspace{1em}A\left\{\frac{1}{{\delta }_0}\left[G_{r+2,3r+2}^{3r+2,1}\left[{\delta }_0❘\begin{array}{c}0,1,{\chi }_1\\ {\chi }_2,0\end{array}\right]-G_{1,0:2,2:r,3r}^{0,1:1,2:3r,0}\times \left[\frac{1}{{\tau }_1},{\delta }_0{\tau }_1❘\begin{array}{c}1\\ -\end{array}❘\begin{array}{c}1,1\\ 1,0\end{array}❘\begin{array}{c}{\chi }_1\\ {\chi }_2\end{array}\right]\right]G_{1,0:2,2:r+1,3r+1}^{0,1:1,2:3r,1}\times \hspace{1em}\left[\frac{1}{{\tau }_1},{\delta }_0{\tau }_1❘\begin{array}{c}2\\ -\end{array}❘\begin{array}{c}1,1\\ 1,0\end{array}❘\begin{array}{c}1,{\chi }_1\\ {\chi }_2,0\end{array}\right]\times {\left({\tau }_1\right)}^2\}\right)+\left(\begin{array}{c}{K}_2-1\\ N_2-1\end{array}\right)K_2\times \sum _{j=0}^{K_2-N_2}\left(\begin{array}{c}{K}_2-N_2\\ j\end{array}\right){\left(-1\right)}^j{\left({\tau }_2\right)}^{-1}(-\exp \left({\tau }_2\right)E_i\left(-{\tau }_2\right)-\hspace{1em}\hspace{1em}\hspace{1em}A\{\frac{1}{{\delta }_0}\left[G_{r+2,3r+2}^{3r+2,1}\hspace{1em}\left[{\delta }_0❘\begin{array}{c}0,1,{\chi }_1\\ {\chi }_2,0\end{array}\right]-G_{1,0:2,2:r,3r}^{0,1:1,2:3r,0}\times \hspace{1em}\left[\frac{1}{{\tau }_2},{\delta }_0{\tau }_2❘\begin{array}{c}1\\ -\end{array}❘\begin{array}{c}1,1\\ 1,0\end{array}❘\begin{array}{c}{\chi }_1\\ {\chi }_2\end{array}\right]]+{\left({\tau }_2\right)}^2{G}_{1,0:2,2:r+1,3r+1}^{0,1:1,2:3r,1}\times \left[\frac{1}{{\tau }_2},{\delta }_0{\tau }_2❘\begin{array}{c}2\\ -\end{array}❘\begin{array}{c}1,1\\ 1,0\end{array}❘\begin{array}{c}1,{\chi }_1\\ {\chi }_2,0\end{array}\right]\}\right)-\left(\begin{array}{c}{K}_1-1\\ N_1-1\end{array}\right)K_1\times \sum _{k=0}^{K_1-N_1}\frac{\left(\begin{array}{c}{K}_1-N_1\\ k\end{array}\right){\left(-1\right)}^k}{{\tau }_1}\left(\begin{array}{c}{K}_2-1\\ N_2-1\end{array}\right)K_2\sum _{j=0}^{K_2-N_2}{\left(-1\right)}^j\frac{\left(\begin{array}{c}{K}_2-N_2\\ j\end{array}\right)}{{\tau }_2}\times (-\exp \left({\tau }_1\right)E_i\left(-{\tau }_1\right)-\exp \left({\tau }_2\right)E_i\left(-{\tau }_2\right)+\exp \left({\tau }_1+{\tau }_2\right)\times E_i\left(-\left({\tau }_1+{\tau }_2\right)\right)-A\left\{\frac{1}{{\delta }_0}\left[G_{r+2,3r+2}^{3r+2,1}\left[{\delta }_0❘\begin{array}{c}0,1,{\chi }_1\\ {\chi }_2,0\end{array}\right]-G_{1,0:2,2:r,3r}^{0,1:1,2:3r,0}\times \left[\frac{1}{{\tau }_1},{\delta }_0{\tau }_1❘\begin{array}{c}1\\ -\end{array}❘\begin{array}{c}1,1\\ 1,0\end{array}❘\begin{array}{c}{\chi }_1\\ {\chi }_2\end{array}\right]-G_{1,0:2,2:r,3r}^{0,1:1,2:3r,0}\left[\frac{1}{{\tau }_2},{\delta }_0{\tau }_2❘\begin{array}{c}1\\ -\end{array}❘\begin{array}{c}1,1\\ 1,0\end{array}❘\begin{array}{c}{\chi }_1\\ {\chi }_2\end{array}\right]+G_{1,0:2,2:r,3r}^{0,1:1,2:3r,0}\left[\frac{1}{\left[{\tau }_1+{\tau }_2\right]},{\delta }_0\left[{\tau }_1+{\tau }_2\right]❘\begin{array}{c}1\\ -\end{array}❘\begin{array}{c}1,1\\ 1,0\end{array}❘\begin{array}{c}{\chi }_1\\ {\chi }_2\end{array}\right]\right]+{\left({\tau }_1\right)}^2\times G_{1,0:2,2:r+1,3r+1}^{0,1:1,2:3r,1}\hspace{1em}\left[\frac{1}{{\tau }_1},{\delta }_0{\tau }_1❘\begin{array}{c}2\\ -\end{array}❘\begin{array}{c}1,1\\ 1,0\end{array}❘\begin{array}{c}1,{\chi }_1\\ {\chi }_2,0\end{array}\right]+{\left({\tau }_2\right)}^2{G}_{1,0:2,2:r+1,3r+1}^{0,1:1,2:3r,1}\left[\frac{1}{{\tau }_2},{\delta }_0{\tau }_2❘\begin{array}{c}2\\ -\end{array}❘\begin{array}{c}1,1\\ 1,0\end{array}❘\begin{array}{c}1,{\chi }_1\\ {\chi }_2,0\end{array}\right]-{\left[{\tau }_1+{\tau }_2\right]}^2\times G_{1,0:2,2:r,3r}^{0,1:1,2:3r,0}\left[\frac{1}{\left[{\tau }_1+{\tau }_2\right]},{\delta }_0\left[{\tau }_1+{\tau }_2\right]❘\begin{array}{c}1\\ -\end{array}❘\begin{array}{c}1,1\\ 1,0\end{array}❘\begin{array}{c}{\chi }_1\\ {\chi }_2\end{array}\right]\}+\frac{1}{{\delta }_0}{G}_{r+2,3r+2}^{3r+2,1}\left[{\delta }_0❘\begin{array}{c}0,1,{\chi }_1\\ {\chi }_2,0\end{array}\right]\right\},& \left(36\right)\end{array}$
where E_i(.) is an exponential integral function, and G[Z₁, Z₂|.|.|.] is the extended generalized bivariate Meijer G-function.

System Performance Metrics (Asymptotic Solutions)

Due to the complexity of the above expressions, approximations of these expressions are required to appropriately evaluate the impact of changes in performance parameters and gain system insight. Detailed derivation of the approximate solutions can be found in the cited references.

The asymptotic outage probability can be written at the high SNR regime as P_out(G_cSNR)^−Gd, where G_c and G_d are the coding gain and diversity order of the system, respectively (see textbook by Simon, M K and Alouini, M-S, “Digital Communication over Fading Channels”, published by John Wiley & Sons, Inc., in 2005, and incorporated herein by reference). G_c represents a horizontal shift in the outage probability and G_d refers to a change in the slope of the outage probability vs. SNR curve.

Consider the case of identical sources' channels (λ_1,r1=λ_2,r1= . . . =λ_K1,r1=λ_u,r1) and identical destinations' channels (λ_r1,1=λ_r1,2= . . . =λ_r2,K2=λ_r2,u). The approximate CDF of each hop separately is determined to calculate the approximate CDF of the e2e SNR.

Regarding the first hop link, using a Taylor series representation of the exponential term in the CDF to simplify and integrate, the CDF is written as

$\begin{array}{cc}{F}_{{\gamma }_{U_{\mathrm{Sel}},R_1}}\left(\gamma \right)\simeq \left(\begin{array}{c}{K}_1-1\\ N_1-1\end{array}\right){K_1\left({\lambda }_{u,r_1}\right)}^{K_1-N_1+1}\frac{{\gamma }^{K_1-N_1+1}}{\left(K_1-N_1+1\right)}.& \left(38\right)\end{array}$

Regarding the second hop link, the CDF is written as

$\begin{array}{cc}{F}_{{\gamma }_{R_1,R_2}}\left(\gamma \right)\simeq {{\rm Y}\left(\frac{\gamma }{{\stackrel{\_}{\gamma }}_{r_1,r_2}}\right)}^{\frac{v}{r}},& \left(40\right)\end{array}$
where γ is constant and is written as

$\begin{array}{cc}{\rm Y}=A\sum _{k=1}^6\frac{\prod _{j=1,j\ne k}^6\Gamma \left(b_j-b_k\right)\Gamma \left(b_k\right)}{\prod _{j=2}^3\Gamma \left(a_j-b_k\right)\Gamma \left(1+b_k\right)}{B}^{{\hspace{0.17em}}^{b}k/2},& \left(43\right)\end{array}$

The third hop link, similar to the first hop link, is simplified as

$\begin{array}{cc}{F}_{{\gamma }_{R_2,D_{\mathrm{Sel}}}}\left(\gamma \right)\simeq \left(\begin{array}{c}{K}_2-1\\ N_2-1\end{array}\right){K_2\left({\lambda }_{r_2,u}\right)}^{K_2-N_2+1}\frac{{\gamma }^{K_2-N_2+1}}{\left(K_2-N_2+1\right)}.& \left(45\right)\end{array}$

To obtain the diversity order and coding gain of the system, the CDF of (16) can be simplified, at high SNR values, to be
F_γD(γ)≅F_γUSel,R1(γ)+F_γR1,R2(γ)+F_γR2,DSel(γ),  (46)

Substituting values into (46), the approximate outage probability, at high SNR values, can be written as

$\begin{array}{cc}{P}_{\mathrm{out}}^{\infty }=\left(\begin{array}{c}{K}_1-1\\ N_1-1\end{array}\right){K_1\left({\stackrel{\_}{\gamma }}_{u,r_1}\right)}^{-\left(K_1-N_1+1\right)}\frac{{\left({\gamma }_{\mathrm{out}}\right)}^{K_1-N_1+1}}{\left(K_1-N_1+1\right)}+{\left(\frac{{{\rm Y}}^{-\frac{r}{v}}}{y_{\mathrm{out}}}{\stackrel{\_}{\gamma }}_{r_{1,}{r}_2}\right)}^{-\frac{v}{r}}+\left(\begin{array}{c}{K}_2-1\\ N_2-1\end{array}\right){K_2\left({\stackrel{\_}{\gamma }}_{r_2,u}\right)}^{-\left(K_2-N_2+1\right)}\times \frac{{\left({\gamma }_{\mathrm{out}}\right)}^{K_2-N_2+1}}{\left(K_2-N_2+1\right)}.& \left(47\right)\end{array}$

From (47), it is observed that the performance of the considered relay network will be dominated by the worst link among the available three links (first RF link, FSO link, second RF link). This domination depends on the parameters of these links. Therefore, the diversity order of the triple-hop mixed RF/FSO/RF relay network with generalized order user scheduling is equal to min(K₁−N₁+1, ν/r, K₂−N₂+1). Based on the value of the diversity order, one of the following three cases represents the overall system performance.

Case 1 One hop is dominant, and the coding gain is written as

$\begin{array}{cc}{G}_c=\{\begin{array}{cc}{\left[\left(\begin{array}{c}{K}_1-1\\ N_1-1\end{array}\right)K_1\frac{{\left({\gamma }_{\mathrm{out}}\right)}^{K_1-N_1+1}}{\left(K_1-N_1+1\right)}\right]}^{-\frac{1}{K_1-N_1+1}},& G_d=K_1-N_1+1,\\ \frac{{{\rm Y}}^{-\frac{r}{v}}}{y_{\mathrm{out}}},& G_d=\frac{v}{r},\\ {\left[\left(\begin{array}{c}{K}_2-1\\ N_2-1\end{array}\right)K_2\frac{{\left({\gamma }_{\mathrm{out}}\right)}^{K_2-N_2+1}}{\left(K_2-N_2+1\right)}\right]}^{-\frac{1}{K_2-N_2+1}},& G_d=K_2-N_2+1\end{array}.& \left(48\right)\end{array}$

Case 2 Two hops are dominant, and the coding gain is written as

$\begin{array}{cc}{G}_c=\{\begin{array}{cc}\frac{1}{2}\{{\left[\left(\begin{array}{c}{K}_1-1\\ N_1-1\end{array}\right)K_1\frac{{\left({\gamma }_{\mathrm{out}}\right)}^{K_1-N_1+1}}{\left(K_1-N_1+1\right)}\right]}^{-\frac{1}{K_1-N_1+1}}+& G_d=K_1-N_1+1=K_2-N_2+1,\\ {\left[\left(\begin{array}{c}{K}_2-1\\ N_2-1\end{array}\right)K_2\frac{{\left({\gamma }_{\mathrm{out}}\right)}^{K_2-N_2+1}}{\left(K_2-N_2+1\right)}\right]}^{-\frac{1}{K_2-N_2+1}}\},& \\ \frac{1}{2}\left\{{\left[\left(\begin{array}{c}{K}_1-1\\ N_1-1\end{array}\right)K_1\frac{{\left({\gamma }_{\mathrm{out}}\right)}^{K_1-N_1+1}}{\left(K_1-N_1+1\right)}\right]}^{-\frac{1}{K_1-N_1+1}}+\frac{{{\rm Y}}^{-\frac{r}{v}}}{{\gamma }_{\mathrm{out}}}\right\},& G_d=K_1-N_1+1=\frac{v}{r},\\ \frac{1}{2}\left\{{\left[\left(\begin{array}{c}{K}_2-1\\ N_2-1\end{array}\right)K_2\frac{{\left({\gamma }_{\mathrm{out}}\right)}^{K_2-N_2+1}}{\left(K_2-N_2+1\right)}\right]}^{-\frac{1}{K_2-N_2+1}}+\frac{{{\rm Y}}^{-\frac{r}{v}}}{{\gamma }_{\mathrm{out}}}\right\},& G_d=K_2-N_1+1=\frac{v}{r},\end{array}& \left(49\right)\end{array}$

Case 3 Three hops have the same diversity order, and so the coding gain is written as

$\begin{array}{cc}{G}_c=\{\begin{array}{cc}\frac{1}{3}\{{\left[\left(\begin{array}{c}{K}_1-1\\ N_1-1\end{array}\right)K_1\frac{{\left({\gamma }_{\mathrm{out}}\right)}^{K_1-N_1+1}}{\left(K_1-N_1+1\right)}\right]}^{-\frac{1}{K_1-N_1+1}}+& \begin{array}{c}\begin{array}{c}{G}_d=K_1-N_1+\\ 1=K_2-N_2+\end{array}\\ 1=\frac{v}{r},\end{array}\\ \frac{{{\rm Y}}^{-\frac{r}{v}}}{{\gamma }_{\mathrm{out}}}+{\left[\left(\begin{array}{c}{K}_2-1\\ N_2-1\end{array}\right)K_2\frac{{\left({\gamma }_{\mathrm{out}}\right)}^{K_2-N_2+1}}{-\left(K_2-N_2+1\right)}\right]}^{-\frac{1}{K_2-N_2+1}}\},& \end{array}& \left(50\right)\end{array}$

System performance, dominated by the weakest link, can be described as: (1) the first hop link (i.e., K₁ and N₁), (2) the second hop link (i.e., ζ², α, β), and (3) the third hop link (i.e., K₂ and N₂). If the diversity orders of two hops are equal and are the minimum, the coding gain of the system equals the average of the coding gains across these two hops. Similarly, if the diversity orders of all three hops are equal, the coding gain of the system equals an average of the coding gains across the three hops.

The above approximate solutions are further used to determine the optimum adaptive power allocation for the transmitting nodes in the system.

The distance between the first hop K₁ sources and relay R₁ is defined as d_s,r1, the distance between the relays R₁ and R₂ is defined as d_r1,r2, and the distance between relay R₂ and the third hop K₂ destinations is defined as d_r2,d. The distance from K₁ sources to K₂ destinations, therefore, is defined as D_tot=d_s,r1+d_r1,r2+d_r2,d. Under a scenario where received power decays with distance, the average value of SNR in the hop between K₁ sources and relay R₁ is expressed as

${\stackrel{\_}{\gamma }}_{s,r_1}=P_{s,r_1}{d}_{s,r_1}^{-\mu },\mathrm{where}{P}_{\mathrm{rs}{,}_2}=\frac{P_{\mu ,r_1}}{K_1}=\frac{E_{s,r_1}}{N_0},$
μ is the path loss exponent and is equal for all hops to a value greater than 1, and N₀ is AWGN power (assumed equal for all hops). Similarly, average value of SNR in the second hop can be expressed as γ_r1,r2=P_r1,r2d_r1,r2^−μ, where

$P_{r_1,r_2}=\frac{E_{r_1,r_2}}{N_0}.$
The average value of SNR in the third hop, between the relay R₂ and destinations K₂, is expressed as

$P_{r_2,d}=\frac{P_{r_2,d}}{K_1}=\frac{E_{r_2,d}}{K_0}.$
The power constraint in this system can, therefore, be written as P_tot=P_s,r1+P_r1,r2+P_r2,d.

The optimal power allocation, minimizing outage probability as a function of the power constraint, is expressed as

$\begin{array}{cc}\left(P_{s,r_1}^{*},P_{r_1,r_2}^{*},P_{r_2,d}^{*}\right)=\arg \underset{\left(P_{s,r_1},P_{r_1,r_2},P_{r_2,d}\right)}{\min }{F}_{\gamma D}\left({\gamma }_{\mathrm{out}}\right),& \left(51\right)\end{array}$

The asymptotic expression for F_γD(γ_out) can be rewritten as

$\begin{array}{cc}{F}_{\gamma D}\left({\gamma }_{\mathrm{out}}\right)\simeq \left(\begin{array}{c}{K}_1-1\\ N_1-1\end{array}\right){K_1\left(\frac{d_{s,r_1}^{\mu }}{P_{s,r_1}}\right)}^{K_1-N_1+1}\frac{{\left({\gamma }_{\mathrm{out}}\right)}^{K_1-N_1+1}}{\left(K_1-N_1+1\right)}+\frac{{\gamma }_{\mathrm{out}}{\mathrm{ABd}}_{r_1,r_2}^{\mu }}{P_{r_1,r_2}}+\left(\begin{array}{c}{K}_2-1\\ N_2-1\end{array}\right)K_2\times {\left(\frac{d_{r_2,d}^{\mu }}{P_{r_2,d}}\right)}^{K_2-N_2+1}\frac{{\left({\gamma }_{\mathrm{out}}\right)}^{K_2-N_2+1}}{\left(K_2-N_2+1\right)}.& \left(52\right)\end{array}$

Using a Lagrangian multipliers method, differentiating, and simplifying to solve for P_s,r1*, P_r1,r2*, and P_r2,d*, the following expressions are achieved for the optimum transmission powers at each hop:

$\begin{array}{cc}{P}_{s,r_1}^{*}=\frac{d_{s,r_1}^{\mu }{\gamma }_{\mathrm{out}}{P}_{\mathrm{tot}}}{{\gamma }_{\mathrm{out}}\left(d_{s,r_1}^{\mu }+d_{r_2,d}^{\mu }\right)+{\left[{\mathrm{ABd}}_{r_1,r_2}^{\mu }{\gamma }_{\mathrm{out}}\right]}^{\frac{1}{2}}},& \left(62\right)\\ P_{r_1,r_2}^{*}=\frac{{\left[{\mathrm{ABd}}_{r_1,r_2}^{\mu }{\gamma }_{\mathrm{out}}\right]}^{\frac{1}{2}}{P}_{\mathrm{tot}}}{{\gamma }_{\mathrm{out}}\left(d_{s,r_1}^{\mu }+d_{r_2,d}^{\mu }\right)+{\left[{\mathrm{ABd}}_{r_1,r_2}^{\mu }{\gamma }_{\mathrm{out}}\right]}^{\frac{1}{2}}},& \left(63\right)\\ P_{r_2,d}^{*}=\frac{d_{r_2,d}^{\mu }{\gamma }_{\mathrm{out}}{P}_{\mathrm{tot}}}{{\gamma }_{\mathrm{out}}\left(d_{s,r_1}^{\mu }+d_{r_2,d}^{\mu }\right)+{\left[{\mathrm{ABd}}_{r_2,d}^{\mu }{\gamma }_{\mathrm{out}}\right]}^{\frac{1}{2}}}.& \left(64\right)\end{array}$

System Evaluation

The accuracy of analytical and asymptotic solutions can be validated via comparison to Monte Carlo simulations.

FIG. 3 is a graphical representation of the impact of the order of selected source at the first hop (N₁) and order of selected destination at the third hop (N₂) on outage probability of the system when N₁=N₂. FIG. 3 demonstrates that under weak turbulence conditions (α=9.708 and β=8.198), or as N₁=N₂ decreases (the quality of the selected source and destination is improved), overall system performance is improved. Under weak turbulence, the RF links in the first and third hop drive system performance, and therefore improvements in SNR at those links dramatically improves system performance. In other words, the diversity order of the system is equal to K₁−N₁+1=K₂−N₂+1. For a fixed number of sources and destinations (K₁=K₂), reducing N_t=N₂ increases the diversity order of the system and enhances system performance.

FIG. 4 is a graphical representation of the outage probability of the system under weak turbulence conditions when the number of sources and destinations is varied but equal (K₁=K₂). Under weak turbulence conditions (α=9.708 and β=8.198) or as K₁=K₂ increases (number of available sources and destinations increases), overall system performance improves. Again, during weak turbulence, the RF links drive system performance and the diversity order is equal to K₁−N₁+1=K₂−N₂+1. In other words, for fixed order of selection of source and destination (N₁=N₂), an increasing K₁=K₂ increases diversity order of the system and improves system performance.

FIG. 5 is a graphical representation of outage probability of the system compared with average SNR at each hop under weak turbulence conditions (α=8.650 and β=7.142) and for different values of the SNR outage threshold γ_out. Two specific cases were evaluated: (1) all links' average SNRs increase with increasing x-axis value and (2) one link's average SNR is fixed. In the case (1) where all SNRs are variable, the performance of the system is not limited and enhances as SNR increases. In the case (2) where one link has a fixed SNR, a noise floor appears in the results and outage probability is dominated by this worst link among the three. From FIG. 5, it is apparent that changes to the outage threshold γ_out affect only the coding gain of the system.

FIG. 6 is a graphical representation of the outage probability of the system compared with order of selected source and destination (N₁=N₂) under weak turbulence conditions (α=8.038 and β=6.525) for variable values of average SNR per hop. As N₁=N₂ increases (as the quality of the selected source and destination decreases), the outage probability of the system increases and system performance degrades.

FIG. 7 is a graphical representation of the impact of the proposed power allocation algorithm under weak turbulence conditions (α=5.662 and β=4.059) and for variable values of outage threshold γ_out. The dashed lines, representing power optimization, improve system performance by decreasing outage probability of the system. Further, by increasing the outage threshold γ_out, system performance is degraded by reducing the coding gain of the system and not the diversity order. In determining power allocation, the total distance (D_tot) between the sources and destinations was assumed to be 1 and divided as D_s,r1=0.3, D_r1,r2=0.3, and D_{r2, d}=0.4.

FIG. 8 is a graphical representation of the impact of pointing error (ζ) on the average symbol error probability of the system under sever atmospheric turbulence conditions (α=4.341 and β=1.309). Under severe atmospheric turbulence conditions, system performance is dominated by performance of the FSO link and its corresponding parameters (α, β, and ζ²). FIG. 8 can be divided into two sets of curves where (1) diversity order is affected by changing ζ and where (2) coding gain is affected by changing ζ. When ζ² is the smallest parameter, G_d of the system is impacted. When ζ is larger than β, changes in ζ affect the G_c of the system and G_d is determined by β.

FIG. 9 is a graphical representation of the impact of FSO detection type (heterodyne or intensity modulation/direct detection) on the average symbol error probability when compared with average SNR per hop under variable atmospheric conditions. Due to minimal sensitivity to thermal noise, heterodyne detection improves system performance while increasing system complexity.

FIG. 10 is a graphical representation of the outage probability of the system when the number of sources K₁ and destinations K₂ and order of selected sources N₁ and destinations N₂ is varied. Further, the additional impact of varied atmospheric turbulence conditions is considered. Under severe atmospheric turbulence conditions, increasing K₁=K₂ does not improve diversity order or coding gain of the system as system performance is dominated by the performance of the FSO link. Under weak turbulence conditions, when the weakest links of the network are the RF links, diversity order and coding gain are dominated by the RF link parameters (K₁, N₁, K₂, and N₂). The diversity order of the system is determined by the minimum value amongst the terms K₁−N₁+1 and K₂−N₂+1. Increasing only one value improves the coding gain, while increasing all values raises the minimum value and improves diversity gain.

FIG. 11 is a graphical representation of the ergodic capacity of the network compared with average SNR per hop under weak atmospheric conditions for varying values of N₁=N₂. By increasing the quality of the source and destination selected, or by decreasing N₁=N₂, system capacity is improved.

FIG. 12 provides a hardware description of the system for wireless network communication according to exemplary embodiments. In FIG. 12, the system for wireless network communication includes a CPU 1200 which performs the processes (user selection, decoding, etc.) described above. The process data and instructions may be stored in memory 1202. These processes and instructions may also be stored on a storage medium disk 1204 such as a hard drive (HDD) or portable storage medium or may be stored remotely. Further, the claimed advancements are not limited by the form of the computer-readable media on which the instructions of the inventive process are stored. For example, the instructions may be stored on CDs, DVDs, in FLASH memory, RAM, ROM, PROM, EPROM, EEPROM, hard disk or any other information processing device with which the system for wireless network communication communicates, such as a server or computer.

Further, the claimed advancements may be provided as a utility application, background daemon, or component of an operating system, or combination thereof, executing in conjunction with CPU 1200 and an operating system such as Microsoft Windows, UNIX, Solaris, LINUX, Apple MAC-OS and other systems known to those skilled in the art.

The hardware elements in order to achieve the system for wireless network communication may be realized by various circuitry elements, known to those skilled in the art. For example, CPU 1200 may be a Xenon or Core processor from Intel of America or an Opteron processor from AMD of America, or may be other processor types that would be recognized by one of ordinary skill in the art. Alternatively, the CPU 1200 may be implemented on an FPGA, ASIC, PLD or using discrete logic circuits, as one of ordinary skill in the art would recognize. Further, CPU 1200 may be implemented as multiple processors cooperatively working in parallel to perform the instructions of the inventive processes described above.

The system for wireless network communication in FIG. 12 also includes a network controller 1206, such as an Intel Ethernet PRO network interface card from Intel Corporation of America, for interfacing with network 1230. As can be appreciated, the network 1230 can be a public network, such as the Internet, or a private network such as an LAN or WAN network, or any combination thereof and can also include PSTN or ISDN sub-networks. The network 1230 can also be wired, such as an Ethernet network, or can be wireless such as a cellular network including EDGE, 3G and 4G wireless cellular systems. The wireless network can also be WiFi, Bluetooth, or any other wireless form of communication that is known.

The system for wireless network communication further includes a display controller 1208, such as a NVIDIA GeForce GTX or Quadro graphics adaptor from NVIDIA Corporation of America for interfacing with display 1210, such as a Hewlett Packard HPL2445w LCD monitor. A general purpose I/O interface 1212 interfaces with a keyboard and/or mouse 1214 as well as a touch screen panel 1216 on or separate from display 1210. General purpose I/O interface also connects to a variety of peripherals 1218 including printers and scanners, such as an OfficeJet or DeskJet from Hewlett Packard.

A sound controller 1220 is also provided in the system for wireless network communication, such as Sound Blaster X-Fi Titanium from Creative, to interface with speakers/microphone 1222 thereby providing sounds and/or music.

The general purpose storage controller 1224 connects the storage medium disk 1204 with communication bus 1226, which may be an ISA, EISA, VESA, PCI, or similar, for interconnecting all of the components of the system for wireless network communication. A description of the general features and functionality of the display 1210, keyboard and/or mouse 1214, as well as the display controller 1208, storage controller 1224, network controller 1206, sound controller 1220, and general purpose I/O interface 1212 is omitted herein for brevity as these features are known.

The exemplary circuit elements described in the context of the present disclosure may be replaced with other elements and structured differently than the examples provided herein. Moreover, circuitry configured to perform features described herein may be implemented in multiple circuit units (e.g., chips), or the features may be combined in circuitry on a single chipset.

Moreover, the present disclosure is not limited to the specific circuit elements described herein, nor is the present disclosure limited to the specific sizing and classification of these elements. For example, the skilled artisan will appreciate that the circuitry described herein may be adapted based on changes on battery sizing and chemistry, or based on the requirements of the intended back-up load to be powered.

The functions and features described herein may also be executed by various distributed components of a system. For example, one or more processors may execute these system functions, wherein the processors are distributed across multiple components communicating in a network. The distributed components may include one or more client and server machines, which may share processing, in addition to various human interface and communication devices (e.g., display monitors, smart phones, tablets, personal digital assistants (PDAs)). The network may be a private network, such as a LAN or WAN, or may be a public network, such as the Internet. Input to the system may be received via direct user input and received remotely either in real-time or as a batch process. Additionally, some implementations may be performed on modules or hardware not identical to those described. Accordingly, other implementations are within the scope that may be claimed.

The above-described hardware description is a non-limiting example of corresponding structure for performing the functionality described herein.

It is notable that for each situation evaluated heretofore, analytical and asymptotic expressions are in match with simulation results.

Obviously, numerous modifications and variations are possible in light of the above teachings. It is therefore to be understood that within the scope of the appended claims, the invention may be practiced otherwise than as specifically described herein.

Thus, the foregoing discussion discloses and describes merely exemplary embodiments of the present invention. As will be understood by those skilled in the art, the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. Accordingly, the disclosure of the present invention is intended to be illustrative, but not limiting of the scope of the invention, as well as other claims. The disclosure, including any readily discernible variants of the teachings herein, defines, in part, the scope of the foregoing claim terminology such that no inventive subject matter is dedicated to the public.

Claims

1. A system for wireless network communication, comprising:

a first plurality of wireless devices at a source configured to transmit or receive communication via radiofrequency;

a second plurality of wireless devices at a destination configured to transmit or receive communication via radiofrequency;

a first relay configured to communicate with a selected one of the first plurality of wireless devices at the source via radiofrequency and communicate with a second relay via free space optical communication, the second relay being configured to communicate with the first relay via free space optical communication and with a selected one of the second plurality of wireless devices at the destination via radiofrequency; and

processing circuitry configured to

select a wireless device at the source with the largest partially modeled source signal-to-noise ratio as the selected one of the first plurality of wireless devices,

select a wireless device at the destination with the largest partially modeled destination signal-to-noise ratio as the selected one of the second plurality of wireless devices, wherein

the partially modeled source signal-to-noise ratio is based on transmitted power, additive white Gaussian noise, and Rayleigh fading model-based channel coefficients, and

the partially modeled destination signal-to-noise ratio is based on transmitted power, additive white Gaussian noise, and Rayleigh fading model-based channel coefficients.

2. The system for wireless network communication of claim 1, wherein the radiofrequency communication of the first plurality of wireless devices at the source is mmWave radiofrequency communication.

3. The system for wireless network communication of claim 1, wherein the radiofrequency communication of the first relay is mmWave radiofrequency communication.

4. The system for wireless network communication of claim 1, wherein the first relay employs a decode and forward scheme.

5. The system for wireless network communication of claim 1, wherein the second relay employs a decode and forward scheme.

6. The system for wireless network communication of claim 1, wherein the radiofrequency communication of the second relay is mmWave radiofrequency communication.

7. The system for wireless network communication of claim 1, wherein the radiofrequency communication of the second plurality of wireless devices at the destination is mmWave radiofrequency communication.

8. The system according to claim 1, wherein the Rayleigh fading model-based channel coefficients are exponentially distributed random variables.

9. The system according to claim 1, wherein the processing circuitry is further configured to

calculate a partially modeled signal-to-noise ratio for the free space optical communication, the calculated partially modeled signal-to-noise ratio being determined in accordance with a Gamma-Gamma fading model.

10. The system according to claim 9, wherein the Gamma-Gamma fading model is a unified Gamma-Gamma fading model including pointing errors.

11. The system according to claim 1, wherein a transmission power of the selected one of the first plurality of wireless devices is calculated as

12. The system according to claim 11, wherein an outage probability of the communicated signal from the selected one of the first plurality of wireless devices to the first relay is calculated as

13. The system according to claim 12, wherein the transmission power of the selected one of the first plurality of wireless devices is determined such that the outage probability, among the first plurality of wireless devices, of the communicated signal is minimized.

14. A method of wireless network communication, comprising:

selecting, by processing circuitry, one of a first plurality of wireless devices at a source, the selected one of the first plurality of wireless devices being a wireless device at the source with the largest partially modeled source signal-to-noise ratio;

transmitting, by the processing circuitry, a communication of the selected one of the first plurality of wireless devices to a first relay via radiofrequency;

receiving and decoding, by the processing circuitry, the transmitted radiofrequency communication from the selected one of the first plurality of wireless devices at the first relay;

forwarding, by the processing circuitry, the decoded communication from the first relay to a second relay via free space optical communication;

receiving and decoding, by the processing circuitry, the forwarded free space optical communication from the first relay at the second relay;

selecting, by the processing circuitry, one of a second plurality of wireless devices at a destination, the selected one of the second plurality of wireless devices being a wireless device at the destination with the largest partially modeled destination signal-to-noise ratio; and

transmitting, by the processing circuitry, the decoded communication from the second relay to the selected one of the second plurality of wireless devices at the destination via radiofrequency, wherein

the partially modeled source signal-to-noise ratio is based on transmitted power, additive white Gaussian noise, and Rayleigh fading model-based channel coefficients, and

the partially modeled destination signal-to-noise ratio is based on transmitted power, additive white Gaussian noise, and Rayleigh fading model-based channel coefficients.

15. The method of wireless network communication of claim 14, wherein the radiofrequency communication of the selected one of the first plurality of wireless devices at the source is mmWave radiofrequency communication.

16. The method of wireless network communication of claim 14, wherein the radiofrequency communication received by the first relay is mmWave radiofrequency.

17. The method of wireless network communication of claim 14, wherein the radiofrequency communication transmitted by the second relay is mmWave radiofrequency.

18. The method of wireless network communication of claim 14, wherein the radiofrequency communication received by the selected one of the second plurality of wireless devices at the destination is mmWave radiofrequency.